de novo 454 assembly

[strob]

Hello,

does anybody has experience with de novo assembly of 454 data? Which tool gave "the best" results?

Many thanks

[mchaisso]

If you don't have access to Newbler, EULER-SR works fairly well (both regular and paired-end): euler-assembler.ucsd.edu

(a plug)
-mark

[kayrt34]

Try CLC Genomics workbench. It's pretty good.

[Raj]

The programs i've used are MIRA and the newest version of Newbler, i think there on version 2, which you should have with your instrument.

MIRA is completely free, runs on linux and is pretty good at assembling more complex assemblies such as those with large number of repeats. The output is in FASTA so you can upload this fasta file with the .sff files generated from the 454 into Newbler. Please note that MIRA takes to close to 24h to assemble a 4.5mb (1.7million reads) genome. Newbler (version 2.0, 64bit took about 30 - 40 min to assembler).

[rociobm]

The programs I´ve used to assemble sequence 454 + Sanger are Newbler, Mira2, CAP3 and Velvet. To the file *.ace, obtained after the assembly, a validation is realized by the program "amosvalidation" (package AMOS). We obtain results very seemed with the Newbler, CAP3 and Mira2. The program Velvet gives us a result completemente differently, very restrictively.

[v_kisand]

does mira2 have a source code available? I would need 32-bit version, older 32bit version of mira does not support 454.

anybody has more tricks etc for using velvet compared to the documentation from its own homepage? concentrating on de novo assembly of normal bacterial genomes...

which ones of assemblers developed for Solid, solexa e.g. short reads is useful for 454 reads? and titanium reads are almost as long as Sanger ones ...hopefully even much longer reads will be reality in few years, is the future switching back to theses tools?

[rociobm]

You can find unload the Mira2 of the following page.

http://www.chevreux.org/projects_mira.html

The 32bit version support 454.

[v_kisand]

From mira readme

6. About 454 data
=================

Please note that the 2.8 line does not officially support assembly of 454
data. These routines are (still) under development, parts of it are already in
included but not compiled in. This is scheduled for the 2.9 development line.

And homepage

Current version: MIRA 2.8.3

The MIRA2 assembler from the stable code tree is available as source code package and as binary packages for 64-bit and 32-bit Linux. The later have been compiled and optimised for processors >= 686 (Intel Pentium IV, AMD Athlon).

PLEASE NOTE: this version is NOT suited for assembly of 454 or Solexa data. Please see the development version below.

And 2.9.xx version is only 64-bit ...
-----------------------------------------

Therefore I do not want to waste time on using 2.8.3

[hlu]

I have used 32 bits for 454 data assembly in the past in 454 FLX, now I completely moved over to 64 bits linux. 32 bits machine just are too painly fully slow to do any thing right now.

Titanium data size is so much bigger, and we are talking about repeat-rich data assembly, 32-bits linux just does not cut it.

Just buy the proper linux hardware, get large memory, and put 64 bits linux on it. It is worth the investment.